Actions
Forum Thread Details
  • State Not Answered
  • Replies 15 replies
  • Subscribers 15 subscribers
  • Views 2278 views
  • Users 0 members are here
Related

Input Device for Disabled - Sense Mouth Open

Former Member

I am interested in building an input device that will provide a computer input for my daughter Liz who has Cerebral Palsy.

Liz has a great deal of involuntary head movement but she can reliably open and close her mouth to actuate a switch.

I would like to provide her with a wireless device that senses her mouth opening and closing to provide input to a single switch scanning communication system.  Years ago I combined a chin strap, a bennie cap and a micro switch to provide that function.  I am looking for ideas!

I have searched many sites looking for devices but all require her head to be held fixed which is impossible for Liz.

Parents
  • 0

    this is not in Any Way meant as an Answer to your/her challenge, just a few ideas.

    I was thinking perhaps something along the lines of an RFID type stick-on tag that would require no power but would activate a Theramin type of receiver, also can she control her breathing? perhaps a small blow-pipe to activate or puncuate or even comunicate using?

    I`m also thinking something along the lines of a Kinect type of non intrusive/intimate contact reader as well, I have no idea How one would code this, and all would certainly need discriminator type filtering for random movements.

    all the best and Gods blessings to you both image

     

    Message was edited by: YT 2095 to add: can you say a little bit more about the Type of modulation of her communication in these movements, does she have Fine controllable skills on top of Large movements or is it Large (Macro) Movements overlaying finer random movements, this May help us elicit cogent data/communications from her and sensor types.

  • 0 in reply to YT2095

    The Kinect approach would utilize facial recognition which I tried 18 years ago.  I built a system using a video camera, frame grabber, image processing with eigenface algo.  It worked if Liz's head was completely still.  Today she can not tolerate her head immobilized.  A second problem is the ambient lighting and getting a frontal image.  The state of the act has certainly improved with huge increases in processing power but the algos are still lacking in reliability.  RFID is a good way to go. No power.  Interrogate the  sensor(s).  Another possibility is to energy harvest from the surrounding area to provide power for the circuits.  Having the sensor directly on her face is the best data collecting site.  The epidermal electronic patch needs to be completely integrated containing either a power source and transmitter or passive RFID type. 

Reply
  • 0 in reply to YT2095

    The Kinect approach would utilize facial recognition which I tried 18 years ago.  I built a system using a video camera, frame grabber, image processing with eigenface algo.  It worked if Liz's head was completely still.  Today she can not tolerate her head immobilized.  A second problem is the ambient lighting and getting a frontal image.  The state of the act has certainly improved with huge increases in processing power but the algos are still lacking in reliability.  RFID is a good way to go. No power.  Interrogate the  sensor(s).  Another possibility is to energy harvest from the surrounding area to provide power for the circuits.  Having the sensor directly on her face is the best data collecting site.  The epidermal electronic patch needs to be completely integrated containing either a power source and transmitter or passive RFID type. 

Children
No Data